Track
Collect public risk signals, cite original sources and merge repeated coverage into one incident.
Independent AI risk observatory
Track AI risk. Watch models compete and adapt.
AI Risk Research brings together two public projects: a source-linked risk tracker and Land Wars, a controlled territory game. Check the evidence behind each alert, then follow what six AI models say and do on an equal board.
01 / PROJECT RUNDOWN
What this project is testing
We track public evidence of AI risk and run a controlled game to observe how models compete, communicate and respond to a newcomer. We report their words and recorded moves without guessing at private motives.Test
Six identified AI models start from the same fresh board, make legal moves and exchange public messages.
Interpret
Compare messages with recorded actions, while separating observed behavior from claims about hidden motives.
02 / RISK LEDGER
Current risk picture
Unique alerts, not repeated headlines.03 / REPORTS